Thread by @DrEricDing

📍ADJUSTED SEVERITY of Omicron vs Delta—Need to discuss *intrinsic severity* versus *observed severity* of #Omicron, which reinfects & evasive against 2-vaccine shots. New 🇬🇧study found that after adjusting for factors, Omicron is only slightly intrinsically milder—2%-12% (red)👇

2) Let’s look at this way… the raw crude rates of Omicron vs Delta doesn’t reflect that Delta isn’t as evasive against vaccines— Omicron **looks** milder because breakthrough (and reinfected) people have higher protection from severe disease. But raw numbers don’t adjust that!

3) from earlier study we see that #Omicron is highly evasive against past immunity from surviving Alpha, Beta, or Delta variant. Little to no neutralization of Omicron among convalescent blood of those people. But they are milder—but this hides the true severity in raw rates.

4) KEY REMINDER— “milder” won’t save us, even if it’s 50% milder — because the exponentially fast doubling time of #Omicron can easily swamp out a 50% lower severity by doubling the case count! And doubling in 2 days!!

https://twitter.com/DrEricDing/status/1473762505092743169

5) My top explanation above is the SIMPLIFIED lay explanation of the Omicron vs Delta severity issue. I recommend people reading this detailed deep dive into the data by @dgurdasani1 — she shares more details if you are curious. Follow her for deeper version insights.

https://twitter.com/dgurdasani1/status/1473764701553283076

6) Also really key… though 3 dose effectiveness against severe disease is good, against infection not so good (and 2 dose poor)—we don’t know what it means for all the countless “mild” #Omicron COVID that we see on impacting actual #LongCovid! Mild infection still ==>Long COVID!

https://twitter.com/dgurdasani1/status/1473767482661998600

7) and there will be BILLIONS of cases of new #Omicron infection in the coming 3 months according to @IHME_UW models. Billions with a B. That is not a typo. The “it’s mild” folks conveniently ignore the bigger picture.

https://twitter.com/drericding/status/1473696761617256453

8) here is a good visual way to see how somehow can merely **look** milder but actually have the same “severe” effects. By enlarging the denominator of the “infected”, the same severe box below is a smaller % of the whole that you detect as cases.

https://twitter.com/nataliexdean/status/1468988174693289994

9) So how does #Omicron add so many cases so that it looks overall milder in the raw %? Well it is highly evasive (for mere infection) against 2 shots of vaccine and evasive against past convalescent immunity— look how they all have little to no neutralization against Omicon 👇

https://twitter.com/drericding/status/1471035675382358018

10) But those with break thoughts and reinfections are definitely milder - the same UK study (top of 🧵) showed 50-70% lower severity among those with a reinfection. This is good. But this also “waters down” the observed severity to artificially *look milder*. Hope it’s clear?

11) But again, in terms of preventing COVID reinfection, past infection (convalescent blood) doesn’t do much against #Omicron.

https://twitter.com/drericding/status/1471038151653314564

12) overall, reinfection is NOT a good thing, even if milder than if never had infection before. Reinfection can lead to #LongCovid too. And #Omicron is king of inducing reinfections while Delta not so much.

https://twitter.com/weselymd/status/1466811571649355779

13) What irks me about the “it’s mild” minimizers out there is that when shown data how hospitalizations are surging in UK, Denmark, or NYC—they just shrug it off with some dismissive remark that “oh they are old / got comorbidities anyway” — AS IF THOSE PEOPLE DO NOT MATTER! 🔥

https://twitter.com/DrEricDing/status/1473762997273337866

14) Reinfections absolutely soaring in UK—nearly all #Omicron. Let this be a lesson for those who tout “natural immunity” or their past infection as if it’s enough. It’s not.

https://twitter.com/chrischirp/status/1474041921249189896

15) Another report indicates it’s just as severe as Delta, but looks milder due to “Simpson’s Paradox” 👇— This is exactly what I talked about above the another dataset. You cannot look at the raw numbers — you need to think like an epidemiologist—adjusting for factors is key.

https://twitter.com/yaneerbaryam/status/1474091037861761026